Adrenocortical carcinoma (ACC) is a highly malignant tumor with an incidence of 1 to 1.6 cases per million per year. It presents with metastatic disease in up to 40% of cases. In advanced or recurrent disease treatment options are limited, and therapies using agents such as mitotane, cisplatin and adriamycin effect a tumor response rate of less than 30%. Thyroid carcinoma is the most common endocrine malignancy, accounting for the majority of deaths from endocrine cancers. Each year in the US, approximately 14,000 new cases of thyroid carcinoma are diagnosed and 1200 patients die from this disease. Conventional therapy consists of surgical resection and radioiodine (131I) therapy. However, for poorly differentiated thyroid carcinomas (PDTCs) and anaplastic carcinomas that do not concentrate iodine, 131I therapy is ineffective. In these patients, therapeutic options are few and largely ineffective. Since treatment options for endocrine cancers are limited, we have begun to examine alternate therapeutic approaches. Transfer of a suicide gene into tumor cells is one strategy for cancer gene therapy. One example of a suicide gene is the herpes simplex virus thymidine kinase gene (HSV-TK). HSV-TK encodes an enzyme absent in mammalian cells that catalyzes the conversion of a nontoxic prodrug, such as ganciclovir (GCV), to its toxic phosphorylated metabolite. The key to successful suicide gene therapy is to restrict gene expression to a defined cell population using specific promoter(s). An approach that restricts expression to a given cell type is applicable to endocrine cancers, because they possess numerous genes with very restricted or specific expression. We began by seeking a molecular marker of adrenocortical tissue. CYP11B1 (11<=-hydroxylase) catalyzes conversion of 11-deoxy-corticosterone and 11-deoxycortisol to corticosterone and cortisol, respectively. Although steroid metabolism is not unique to the adrenal gland, biochemical studies have suggested CYP11B1 activity is confined to the adrenal cortex. The specificity of CYP11B1 expression was confirmed using 30 normal tissues. Expression was found only in normal adrenal tissue at an arbitrary level of 10, with a level of 0.1 in testes and ovary and undetectable expression in 27 other normal tissues. More importantly, expression was documented in 32/32 ACCs, with very high levels of expression in about half. Expression was not detected in other tumors. This expression profile makes this gene an attractive choice for a target specific therapy. Since expression of CYP11B1 was highly specific for adrenal tissue and ACCs, the putative CYP11B1 promoter was cloned proximal to an HSV-TK gene. In ACC cells a plasmid containing the HSV-TK gene under the control of the CYP11B1 promoter resulted in HSV-TK expression and enhanced sensitivity to GCV. These findings suggested the CYP11B1 promoter could be used to target ACC cells in suicide gene therapy. Although CYP11B1 promoter activity was preferentially observed in ACC cells, we set out to enhance it while maintaining its specificity for adrenal cells. Placing a putative enhancer element from the cholesterol side-chain cleavage (P450scc) gene upstream of the CYP11B1 promoter (SCC/11<= construct) enhanced its activity and specificity for ACC cells. A second construct using the P450scc promoter and enhancer (SCC/SCC) was also shown to have a high adrenal preference. Using the SCC/11<= and SCC/SCC vectors we constructed suicide HSV-TK adenoviruses and demonstrated their utility in vitro. GCV treatment demonstrated at least a 100-fold greater sensitivity of H295 cells at all concentrations of GCV, and exclusive cytotoxicity at clinically achievable concentrations. Encouraged by our success using adrenal specific promoters, and to determine whether expression could be modulated in other endocrine tumors, we carried out similar studies in thyroid cancer cells. These studies have shown that the thyroglobulin (Tg) promoter can direct specific expression of HSV-TK in thyroid cancer cells. Furthermore, a putative enhancer element for the Tg gene, augmented the activity and specificity of the Tg. In transient or stably-transfected thyroid carcinoma cells, treatment with the histone deacetylase (HDAC) inhibitors, depsipeptide (FR-901228, FK228) or sodium butyrate, alone or in combination with 8-Br-cAMP, resulted in further enhancement. Similar results were seen in two follicular and two anaplastic carcinoma cell lines, suggesting efficacy in the full spectrum of thyroid malignancies. The drug-mediated enhancement was not found in the non-thyroid cells. Enhancement of GCV sensitivity of as much as 100,000- fold was achieved despite low transfection efficiencies. Our studies have demonstrated that depsipeptide increases the efficiency of adenoviral transgene expression in vitro in cancer cells, in hematopoietic cells and in human umbilical vein endothelial cells (HUVEC) and may be useful in cancer gene therapy. Treatment with minimally cytotoxic doses of depsipeptide increases CAR and av integrin RNA preferentially in cancer cells. Ongoing experiments have demonstrated that depsipeptide can increase the efficiency of adenovirus infection in vivo and this occurs in a variety of models. Finally, surprised by the change following the addition of HDAC inhibitors, we examined the ability of depsipeptide to modulate expression of thyroid specific genes. Two follicular (FTC 132 and FTC 136) and two anaplastic thyroid carcinoma cell lines (SW 1736 and KAT-4) were treated with a sub-cytotoxic concentration of depsipeptide (1 ng/ml). After three days, Tg and Na+/I- symporter (NIS).mRNA levels approached those of a normal thyroid. 125I accumulations indicated a functional NIS was induced. These in vitro results suggest depsipeptide or another HDAC inhibitor might be used clinically in thyroid carcinomas that do not to trap iodine, as an adjunct to radioiodine therapy a strategy currently under investigation in the clinic It is thought that CAR is important in adenovirus selectivity. In contrast to other vectors adenovirus infect a variety of organs with high efficiency, with transducibility depending to a certain extent on CAR expression. The lack of CAR expression limits the efficacy of adenovirus-mediated gene transfer for cancer cells and increased CAR levels increase cancer cell susceptibility to adenovirus mediated gene transfer. CAR expression has been demonstrated in human and mouse liver, and is likely responsible for the liver toxicity of Ad-TK gene therapy, especially when non-specific promoters are used. We will optimize conditions for drug dose and infection time relative to drug treatment in order to obtain maximum transgene expression in the adenovirus infected melanoma xenograft-bearing mice. Finally with both poorly differentiated thyroid carcinomas and anaplastic thyroid carcinomas, therapeutic options are limited and largely unsuccessful. Their inability to trap iodine is thought to be a consequence of a loss of expression of the Na+/I- symporter (NIS). Our results suggest HDAC inhibitors up-regulate NIS transcription. Our clinical experience with depsipeptide has found it to be well tolerated, and the levels achieved greatly exceed those that modulated expression of the thyroid genes. A phase one trial to test this observation is ongoing.